CHAPTER 9. ROLE OF THE PHYSICIAN IN SMOKING CESSATION 451 CONTENTS Introduction Patient Groups General Practice Patients Pregnant Patients Patients With Pulmonary Disease Patients With Cardiac Disease The Use of Nicotine Chewing Gum Discussion and Synthesis Methodological Considerations Trends in the Literature Patient Variables Related to Abstinence Physician Variables Related to Effectiveness Conclusions Recommendations for Physicians Future Research Summary and Conclusions References 453 Introduction Although a variety of health care providers have attempted to change the smoking behavior of the groups with whom they work (USDHEW 1979), most of the research in this area, and this review, is confined to patient populations or patient groups who provide opportunities for physician intervention. The nature and extent of the relationship between patient and physician enhances the oppor- tunity for long-term behavior change. Major international studies on utilization of health care reveal that 70 percent or more of North Americans see a physician at least once a year (Kahn and White 1976; National Center for Health Services Research 1983; Pacific Mutual 1978; National Center for Health Statistics 1982). Given this frequency of contact between smokers and their physicians, some 38 million of the 54 million adults in the United States who smoke could be reached annually with a smoking cessation message. Even if only 5 to 10 percent quit on a long-term basis, the potential impact of such contact is enormous. A recent study comparing free medical care to insurance plans requiring shared cost by participants did not show a beneficial impact on smoking (or other health habits associated with coronary heart disease and some types of cancer) from the average of one to two more encounters per year for several years (Brook et al. 1983). The authors comment that "these health habits, especially smoking, were at levels at which substantial health benefit from behavior change was possible" (p. 1432). Thus, physician contact alone does not increase smoking behavior change. Rather, a mix of physician, motivation, educational, and training efforts are doubtless called for. Many techniques have been described in the literature to assist physicians in treating cigarette smoking in their patients (Allaire 1983; Best 1978; Bohm and Powell 1982; Danaher et al. 1980; Fowler 1983; Hochbaum 1975; Hymowitz 1977; Indyke and Ellis 1980; Luban-Plozza 19'77; Pechacek and Grimm 1983; Pechacek and McAlister 1980; Pomerleau 1976; Rose 1975/76; Rosser 1977; Russell 1971; Seeker-Walker and Flynn 1983; Sherin 1982; Shipley and 3rleans 1982; Windsor et al. 1979). These range from supplying nformation about smoking and health with advice to quit smoking ;o implementing complex behavior modification techniques with -outine monitoring and long-term followups. Lichtenstein and Danaher (1978) have described a hypothetical node1 for the various roles that the physician can perform. iccording to their formulation, the physician can "(1) act as a model If a healthy lifestyle by not smoking, (2) provide information larifying the risks associated with smoking and the risk reduction if he patient stops, (3) encourage abstinence by direct advice and uggestions, (4) refer the patient to a smoking cessation program, nd (5) prescribe and follow up the use of specific cessation and 455 maintenance strategies in his or her own office management" (p. 233). This scheme is a hierarchical one, with each role subsuming the behavior of the ones that precede it. Other roles such as political lobbyist and researcher are related only indirectly to patient care (Rosen and Ashley 1978). In addition to advising their patients to quit, an overwhelming majority of physicians have quit smoking; the prevalence of smoking among physicians has most recently been estimated at 10 percent or less in the United States, considerably below that of the general population (Enstrom 1983; Fletcher and Doll 1969; Garfinkel 1976; USDHEW 1976; Sachs 1983). Physicians are, therefore, carrying out their role as exemplars. In a major national survey, over 90 percent agreed that it was their responsibility to set a good example for patients by not smoking cigarettes (USDHEW 1976). With regard to other roles, the majority of reports (both research and advisory) indicate that physicians usually function as information providers and advice givers. However, evaluations of treatment procedures that can be used for referral or in-depth treatment have also been carried out. It is expected that as results become known and more referral agencies are availa.ble, more physicians will be expanding their roles. The literature on rates at which physicians advise patients to quit smoking shows a disparity between physician estimates and patient reports. Over time, the proportion of physicians recommending cessation has increased dramatically. A mid-1960s survey revealed that 38 percent of physicians claimed that they advised "all" or "almost all" (95 to 100 percent) of their patients who did not have smoking-related disorders ,to quit or cut down (Green and Horn 1968). Eighty-eight percent of physicians claimed they gave this advice to patients with lung and pulmonary conditions. In 1979,85 to 92 percent of physicians participating in the evaluation of a quit smoking kit said they had spoken to smoking patients in the past few weeks, advising quitting t,o 6 to 7 out of the last 10 smoking patients seen (American Cancer Society 1981). In a 1981 survey of primary care practitioners in Massachusetts, 90 percent of all physicians who responded said they routinely asked about smoking; however, only 58 percent felt "very prepared" to counsel patients, and a mere 3 percent felt they were currently "very successful" in helping patients to change their smoking behavior (Wechsler et al. 1983). Ninety-eight percent of a Canadian sample of primary care physi- cians surveyed in late 1981-1982 reported advising patients who smoke to stop, with 45 percent claiming some success (Battista 1983; Battista and Spitzer 1983). There is evidence that smoking physi- cians feel less comfortable in dispensing advice to quit smoking and, therefore, do it less forcefully (American Cancer Society 1981). 456 The majority of persons who smoke feel that physician advice to quit or cut down on smoking would be influential (American Cancer Society 1977; Pacific Mutual 1978). In a 1978 survey of the public, doctor's advice was perceived to be the most effective means of prompting cessation or reduction among six alternatives considered, the other five being prohibition of smoking at work and in public places; urging by children, spouse, or relatives; higher taxes on tobacco; antismoking informational campaigns at work; and anti- smoking advertising on television (Pacific Mutual 1978). In this survey, 76 percent of smokers reported that doctor's advice would be "very" or "somewhat" effective in this regard. Given this general level of enthusiasm and confidence in physician-delivered messages, actual rates of reported advice are quite low. In the survey just reported, only 8 percent of former smokers spontaneously mentioned a doctor's recommendation as a cause of their cessation, although 51 percent cited health reasons (Pacific Mutual 1978). In the 1975 Adult Use of Tobacco survey, a full 64.6 percent of male and 60.8 percent of female current smokers claimed they had never received advice from any doctor about quitting, cutting down, or continuing smoking (USPHS 1976). About 20 percent of current smokers had been advised to quit. Combining advice to quit or cut down, the percentage rose to approximately 35 percent. A somewhat lower estimate of physician advice was obtained from a nationwide study of approxi- mately 8,000 people (Stewart et al. 1979). Advice to quit or cut down was reported by 22.4 percent, and lack of advice by 77.6 percent. However, patient recall for the details of a physician visit may be flawed. In one study, almost complete recall of cessation advice was reported 1 year later (Mausner 19701, but in a second study only 50 percent of patients recalled cessation advice 2 months after it was given (Rose and Udechuku 1971). It seems quite likely that physicians do offer varying degrees of advice and guidance to their patients (Fowler and Jamrozik 1983; Wechsler et al. 19831, and in view of the decrease in social acceptability given to the smoker, more physicians will be spending more of their time and energy in this way in the future. A growing number of editorials in medical journals have been devoted to the importance of primary prevention and to motivating physicians to this task (Check 1979; Yankauer 1983). This chapter reviews and summarizes studies of smoking cessation in various groups of patients, with a special focus on physician intervention. Four classes of patients are considered: general prac- tice, obstetric, pulmonary disease, and cardiovascular disease. Re- views of this literature show a positive relationship between severity of disease and the likelihood of quitting smoking (USDHEW 1979, 1980; Lichtenstein and Danaher 1978; Pederson 1982). However, Pederson cautions that a causal interpretation of this relationship 480-144 0 - 85 - 16 may not be warranted, as physician involvement may be greater and the effect of physician advice more salient or intense with sicker patients. In addition, a section on research using nicotine chewing gum as a treatment is included. Suggestions regarding future research and treatment are also presented. Patient Groups General Practice Patients Unlike the physician whose practice involves mainly patients with pulmonary or cardiac disease, a large proportion of the general practitioner's time may be spent in lifestyle modification of a preventive nature with patients who are not experiencing smoking- related problems. It may be that compliance among these patients is dependent upon diagnosis, but no available studies have related the reason for the office visit to success or failure in quitting, although most counseling is said to take place during regular checkups or during visits for respiratory problems, much less often than during visits for unrelated major medical problems or minor problems (Battista 1983). Nine studies have dealt specifically with general practice pa- tients. Mausner et al. (1968) followed 157 smoking patients of two physicians sharing an office. One physician advised all the smokers in his practice who came to his office over a 2-month period to quit (n=121). Patients were told that smoking was harmful, and were given written information on quitting techniques as well as lobeline, a nicotine substitute. The other physician made no special mention of smoking (n= 36 patients). At a 6-month followup, 33 percent of those who were told to quit had reduced the amount they smoked, compared with 9 percent in the group without such cessation advice. Reduction was defined as a decrease of at least 10 cigarettes per day. There was no validation of self-report. The factors found to be related to decrement in smoking were higher initial consumption and number of pack-years; a marginally significant relationship with being male was noted, and for both sexes it was the heavier smokers who changed. Porter and McCullough (1972) compared the smoking behavior of 101 randomly selected patients who were counseled by one general practitioner about their smoking with 90 patients who were not counseled. Counseling consisted of advice, discussion, and a leaflet. There was no significant difference in quit rate after 6 months between the two groups: 2.5 percent in the counseled group and 4.4 percent in the not counseled group quit. No validation was per- formed. Handel (1973) followed for 1 year a group of 100 patients whom she had advised in l- to 7-minute messages to quit smoking. The advice 458 was followed by 38 percent of the men and 11 percent of the women. Eighteen percent of the remaining male smokers and 22 percent of the female smokers reported reducing consumption by more than 50 percent. No control group was included in this study. Pincherle and Wright (1970) reported a smoking intervention in a clinic that provided health examinations of business executives on an annual or biannual basis. Physicians were encouraged to deliver a strong antismoking message, and a booklet was made available. Results varied among the 10 participating physicians; between 17 and 35 percent of the 1,493 smokers seen at a followup visit at approximately 18 months had stopped smoking cigarettes or had reduced their smoking more than 30 percent, There were no controls or validation of self-report. The doctor's own past or present smoking habits only partially accounted for the variation in success rates (cf. American Cancer Society 1981). The quit rate of 19 percent reported by Richmond (1977) in a similar setting is consistent with their findings. In preliminary findings, Rosser (1979) reported that 10 percent of smokers counseled by family physicians about cardiovas- cular risk reduction report smoking cessation 1 or 2 years later. In a large-scale study of 2,138 patients of 28 London physicians in five practices, Russell et al. (1979) assessed the effectiveness of physician smoking advice in comparison with no advice. Assignment to group was by day of attendance at practice. Four groups were used: a nonintervention control, a questionnaire-only control, an advice-only group, and an advice group receiving a two-page pamphlet and a warning of subsequent followup. Advice was delivered in the physician's own style in a l- or 2-minute message. At l-year followup, the overall quit rate was 14.4 percent-respectively for each group, 10.3, 14.0, 16.7, and 19.1 percent. The percentages of patients who stopped within 1 month of the initial visit and who were still abstinent at followup were 0.3, 1.6, 3.3, and 5.1 percent, respectively. These results were statistically significant, indicating that advice to quit was effective and enhanced by written material and information about a subsequent followup. The major effect was to increase motivation in terms of the percentage of patients in each group attempting to quit but not the success rate of quit attempts, and to reduce relapse at the l-year point compared with the initial l- month assessment. One can interpret this as due to the limited scope of advice, focused on health education, and not to quitting skills. Quit rates differed markedly among physicians and were inversely related to the patient's initial consumption. Validation of the verbal report of abstinence on a very small subsample of patients, using a measure of nicotine concentration in saliva, revealed a low deception rate (7 percent), which may have been unreliable, owing to patient selection methods. 459 In an attempt to replicate findings of Russell et al. (1979) in a Canadian sample, Stewart and Rosser (1982) randomly assigned 691 patients to one of three groups: control, advice, and advice plus pamphlet. There were no differences between the groups; only 3 to 4 percent of patients had stopped smoking at the 5-month followup and were still abstinent at 1 year. At that later followup, the overall success rate was 11.7 percent; no objective measure of smoking status was included. The researchers note that the control group had a higher rate of long-term quitting (3.1 percent) than in the Russell group's 1979 study (0.3 and 16 percent). A second study by Russell et al. (1983131 enrolled a sample of 1,938 cigarette smokers, aged 16 and older, who visited 34 general practitioners in six group pra.ctices in Kent and London in November 1980. All smokers were included and were assigned in balanced design by week of attendance to one of three groups; nonintervention controls, 1 to 2 minutes of advice in the physician's own style plus booklet and warning of followup, and similar advice plus booklet plus offer of a nicotine gum prescription. A questionnaire was mailed and a personal followup was performed after 4 months and 1 year. Patients who did not provide adequate data at both points were counted as smokers. Two-thirds of those who claimed to have quit at each time point were checked by measurement of expired air carbon monoxide. At 1 year, self-reported quit smoking rates in the three groups were 13.4, 10.8, and 16.2 percent (p ~0.021, respectively. For those patients not smoking at 4 months and at 1 year, the cessation rates were 6.0, 6.4, and 11.9 percent, respectively (p ~0.02). After correction for those who refused or failed chemical validation (22 percent) and for those who switched from cigarettes to pipes or cigars, the cessation rates were 3.9, 4.1, and 8.8 percent (p